Search

A scale-up methodology was applied to the model previously developed for the industrial production of vinegar in an aerated-stirred pilot fermenter. The response surfaces generated by the model demonstrated that aerated mechanical power input, superficial air velocity, temperature, hydrostatic pressure and concentrations of compounds must be kept in the industrial-scale fermenters with the same geometry dimensions as the pilot fermenter. These variables maintain the values of kLa, C O 2 , L ∗ and OTRmax and asses the same effects of hydrodynamics and oxygen transfer, allowing the values of the process variables of the industrial-scale fermenters to be scaled down to the values of the pilot fermenter, which are required for applying the models. On these bases, a Simulation Environment for Acetification Processes (SEAP) was developed to design industrial processes on any scale and to predict the behaviour of fermentations and the costs of vinegar production. Finally, an example of the design and optimisation of a fermentation process with SEAP is reported, applying an optimisation methodology based on response surfaces and genetic algorithms using a desirability function as the response.

The production of vinegar in an aerated-stirred reactor was modelled using a hybrid approach. The growth rate was modelled according to the total biomass specific growth rate, μ g ν , which represents the growth of an active biophase and the fraction of cells which are growing. A quadratic model based on the process variables, i.e. hydrostatic pressure, ethanol concentration, aeration, agitation and temperature, was developed for μ g ν and the substrate and product kinetics were related mechanistically to the growth rate, including the kinetics of acetoin and ethyl acetate production. This model was coupled with the models for oxygen transfer and gas–liquid transfer, reported in a previous study, and the prediction capacity of the system of equations was validated by simulating semi-continuous and continuous real processes developed in a pilot fermentor.

The transfer of oxygen from aeration flow to the fermentation medium and the transfers of ethanol, ethyl acetate, acetoin, acetic acid and water from the fermentation medium to aeration flow were modelled for the industrial production of vinegar. The study was developed in a pilot fermentor, working with continuous processes and industrial fermentation mediums. The cores of the proposed models were the semi-empiric correlations for the volumetric transfer coefficients kLa for oxygen transfer and kA,Ga for the liquid–gas transfers. The main novelty was the introduction of the effect of pressure on the fermentation medium, together with the expression proposed for kA,Ga, with the same structure as the expression for kLa. A genetic algorithm was developed to calculate the parameters of the models, with a four composed desirability function as the evaluation function. The optimised sets of parameters provided models with good predictive ability. The models can be applied for any fermentor with the same agitation and aeration system design and the same geometry.

The basic tool to simulate the evolution of a bioprocess, such as the industrial process of acetic fermentation, is the kinetic model. The model must be simple and have a high predictive ability to give results capable of explaining the real behaviour. The difficulties in the kinetic modelling of biological processes are mainly related to the description of the bacterial growth. In this paper a genetic algorithm is designed to obtain a set of kinetic parameters for the specific growth rate that enables the model to explain the industrial fermentation. Only acetic concentration data from industrial fermentators are required. A four-factor desirability function works properly as the response to maximize. The model optimized can explain not only the behaviour of the industrial batch process studied, but also the experimental results obtained with a pilot fermentator working continuously with and without cell recycling.

The modeling and optimization of industrial processes requires an intensive study of the factors involved. In this work, a continuous pilot system for studying the industrial process of acetic fermentation is developed. A Doehlert design is applied to the five variables involved in the pilot process. This experimental design allows reduction of the experimental burden and the maximum amount of information to be obtained, studying the factors at different levels depending on their significance. The experimental system provides a robust measure of the specific growth rate and the rates of substrates consumption and acetic acid production, related to the flow of effluent stream evaluated in the steady state. The results demonstrate the growth-associated kinetics of substrates and product, and the yield factors are calculated with low values of variances for the coefficients, i.e., within the range 1-11%. The specific growth rate suits the quadratic model proposed. The response surfaces generated by the model are applied to explain the behavior of the bacterial growth and, therefore, the effects of the process variables studied over the acetic acid production. Very low levels of ethanol or oxygen make the acetification rate decrease, and a saturation effect with high levels of ethanol or oxygen is also deduced. The effects of the aeration rate, agitation, and overpressure suggest a kind of inhibition of the acetic acid production caused by the oxygen that has not been practically studied before. The temperature strengthens the inhibitory effect of the ethanol and the oxygen. The conclusions of this work consolidate the structure of a hybrid model for the acetic fermentation.

The most important kinetic models developed for acetic fermentation were evaluated to study their ability to explain the behavior of the industrial process of acetification. Each model was introduced into a simulation environment capable of replicating the conditions of the industrial plant. In this paper, it is proven that these models are not suitable to predict the evolution of the industrial fermentation by the comparison of the simulation results with an average sequence calculated from the industrial data. Therefore, a new kinetic model for the industrial acetic fermentation was developed. The kinetic parameters of the model were optimized by a specifically designed genetic algorithm. Only the representative sequence of industrial concentrations of acetic acid was required. The main novelty of the algorithm is the four-composed desirability function that works properly as the response to maximize. The new model developed is capable of explaining the behavior of the industrial process. The predictive ability of the model has been compared with that of the other models studied.

The application of near infrared(NIR) spectroscopy to measure simultaneously the concentration both of substrates and products of acetic fermentation, i.e. ethanol, acetic acid and biomass concentration, is reported. All these three parameters are closely involved in the kinetics of the process. Monitoring of these variables is a valuable tool to model, optimise and control industrial processes. Furthermore, the NIR technique is also applied to measure the concentrations of ethyl acetate and acetoin, two compounds of decisive relevance for the sensory characterisation of commercial vinegar. The monitoring of these compounds is fundamental to the global optimisation of the process both in economic and sensorial terms. The analysed samples were taken directly from the industrial fermentors of a vinegar producer in La Rioja and their NIR spectra were measured in order to serve as the basis for the subsequent development of separate PLS regression models for each studied response. Several pre-processing methods, such as mean centring, autoscaling, derivation, SNV, MSC and OSC were applied to the data set, testing and comparing the quality of the respective calibration models constructed. The simplest model that provided the best results in terms of predictive ability was selected for each parameter. Thus, the NIR instrument was integrated into a pilot fermentation system to monitor the concentration of ethanol, acetic acid, ethyl acetate, acetoin and biomass (critical parameters to model and optimise the industrial process) by using the proposed regression models.